On Tue, Nov 09, 2021 at 08:57:48PM +0100, Bernhard Reutner-Fischer wrote:
> On Tue, 9 Nov 2021 20:47:02 +0100
> Bernhard Reutner-Fischer <rep.dot....@gmail.com> wrote:
>
> > On Tue, 9 Nov 2021 14:17:14 -0500
> > Marek Polacek <pola...@redhat.com> wrote:
> >
> > > + if (!valid_p (vendor_start, vendor_len)
> > > + || !valid_p (attr_start, attr_len))
> > > + {
> > > + error ("wrong argument to ignored attributes");
> > > + continue;
> > > + }
> > > + canonicalize_attr_name (vendor_start, vendor_len);
> > > + /* We perform all this hijinks so that we don't have to copy OPT.
> > > */
> > > + tree vendor_id = get_identifier_with_length (vendor_start,
> > > vendor_len);
> >
> > [[____::attry]] void f17y(); /* { dg-warning "ignored" } */
> >
> > so i'd maybe put the
> > || !canonicalize_attr_name (vendor_start, vendor_len)
> > in the condition above?
>
> Well no. Guess vendors and attribute just consisting of underscores
> will not be used that often anyway. I'd have required at least one
> alpha or at least alnum but whatever.
> thanks,
I see that -Wno-attributes=vendor::____ or -Wno-attributes=____::____
crashes, so I need to send yet another version to handle this.
*sigh*
Marek
